Overexpression of PVR and PD-L1 and Its Association with Prognosis
Total Page:16
File Type:pdf, Size:1020Kb
www.nature.com/scientificreports OPEN Overexpression of PVR and PD‑L1 and its association with prognosis in surgically resected squamous cell lung carcinoma Jii Bum Lee1,2, Min Hee Hong1, Seong Yong Park3, Sehyun Chae4, Daehee Hwang5, Sang‑Jun Ha6, Hyo Sup Shim7* & Hye Ryun Kim1* Targeting T‑Cell Immunoreceptor with Ig and ITIM domain‑poliovirus receptor (PVR) pathway is a potential therapeutic strategy in lung cancer. We analyzed the expression of PVR and programmed death ligand‑1 (PD‑L1) in surgically resected squamous cell lung carcinoma (SQCC) and determined its prognostic signifcance. We collected archival surgical specimens and data of 259 patients with SQCC at Yonsei Cancer Center (1998–2020). Analysis of variance was used to analyze the correlations between PVR and PD‑L1 expression and patient characteristics. Kaplan–Meier curves were used to estimate recurrence‑free survival (RFS) and overall survival (OS). Most patients were male (93%); the majority were diagnosed with stage 1 (47%), followed by stage 2 (29%) and stage 3 (21%). Overexpression of PVR resulted in a signifcantly shorter median RFS and OS (P = 0.01). PD‑L1 expression was not signifcant in terms of prognosis. Patients were subdivided into four groups based on low and high PVR and PD‑L1 expression. Those expressing high levels of PVR and PD‑L1 had the shortest RFS (P = 0.03). PVR overexpression is associated with a poor prognosis in surgically resected SQCC. Inhibition of PVR as well as PD‑L1 may help overcome the lack of response to immune checkpoint monotherapy. Squamous cell lung carcinoma (SQCC) accounts for 20–30% of all non-small cell lung cancers (NSCLCs)1. With distinct clinicopathological features, such as a lack of targetable mutations, higher incidence in older individuals, and advanced or metastatic disease at diagnosis, patients with SQCC have a shorter lifespan than those with other NSCLC subtypes2–4. Despite surgical resection, many patients with NSCLC experience relapse5. In surgically resected NSCLC (pathological stages IA–IIB), the 5-year survival rate is approximately 40–70%, but drops to 13% and 5% for stage IIIA and IIIB disease, respectively 6. Prior to immunotherapy, the frst-line treatment options for advanced SQCC were limited to platinum- doublet chemotherapy, with a median survival of 8–11 months 7. Te advent of pembrolizumab, a programmed cell death protein-1 (PD-1) inhibitor, revolutionized the frst-line treatment options for both patients with squamous and non-squamous NSCLC with programmed death ligand-1 (PD-L1) expression levels of ≥ 50%8. However, only 30% of patients with advanced NSCLC exhibit high levels of PD-L1 expression and are eligible for treatment with pembrolizumab8,9. Recently, adding pembrolizumab to chemotherapy resulted in a signifcant survival advantage for patients with advanced SQCC in frst-line settings, regardless of PD-L1 expression 10. Although the PD-L1 tumor propor- tion score is currently a biomarker for frst-line treatment using pembrolizumab, the predictive role of PD-L1 in combination treatment needs further validation since patients across all categories of PD-L1 expression showed improved outcomes with combination treatment. Other combination treatments with immune checkpoint inhibi- tors—such as ipilimumab, a monoclonal antibody against cytotoxic T-lymphocyte-associated protein-4—and 1Division of Medical Oncology, Department of Internal Medicine, Yonsei Cancer Center, Yonsei University College of Medicine, 50-1 Yonsei-ro, Seodaemun-gu, Seoul, Republic of Korea. 2Division of Hemato-oncology, Wonju Severance Christian Hospital, Yonsei University College of Medicine, Wonju, Republic of Korea. 3Department of Thoracic and Cardiovascular Surgery, Yonsei University College of Medicine, Seoul, Korea. 4Korea Brain Bank, Korea Brain Research Institute, Daegu, Korea. 5Department of Biological Sciences, Seoul National University, Seoul, Korea. 6Department of Biochemistry, College of Life Science & Biotechnology, Yonsei University, Seoul, Korea. 7Department of Pathology, Yonsei University College of Medicine, 50-1 Yonsei-ro, Seodaemun-gu, Seoul, Republic of Korea. *email: [email protected]; [email protected] Scientifc Reports | (2021) 11:8551 | https://doi.org/10.1038/s41598-021-87624-x 1 Vol.:(0123456789) www.nature.com/scientificreports/ Figure 1. Programmed death ligand-1 (PD-L1) and poliovirus receptor (PVR) expression analysis using immunohistochemistry with ×200 magnifcation. (a) PD-L1hi/PVRlo, (b) PD-L1lo/PVRlo, (c) PD-L1hi/PVRhi, and (d) PD-L1lo/PVRhi. chemotherapy did not improve overall survival (OS) in patients with metastatic squamous or non-squamous NSCLC11. Other checkpoint receptor blockers that modulate immune cell activation are currently under investigation since a signifcant proportion of patients do not respond to immunotherapy 12. Among checkpoint blockers, T-cell immunoglobulin and ITIM domain (TIGIT) binds to poliovirus receptor (PVR). TIGIT is enriched on T-cells, such as regulatory T-cells, type 1 regulatory T-cells, memory T-cells, exhausted CD8 + T-cells, natural killer T-cells, and natural killer cells13,14. Mouse models have shown that TIGIT/PVR binding induces immune evasion of tumor cells, one of the key hallmarks of cancer15,16. Levels of PVR are usually low (or PVR is not expressed) in normal tissues, but are high in tumor cells. PVR expression is associated with the invasion, migration, and proliferation of tumor cells17. Overexpression of PVR is observed in multiple malignancies, and tumors express- ing high levels of PVR are also associated with poor prognosis18–20. Recently, several clinical trials have evaluated the efcacy of blocking the TIGIT/PVR axis in solid cancers such as NSCLC21. Tiragolumab, an anti-TIGIT antibody, in combination with atezolizumab, an anti-PD-L1 agent, showed a promising treatment response and a manageable toxicity profle in frst-line NSCLC treatment. Compared to monotherapy (atezolizumab), combination therapy improved the objective response rate (31.3% vs. 16.2%) and median progression-free survival (5.4 vs. 3.6 months). Tese results strongly suggest that the TIGIT/PVR axis may be a clinically useful target for treating patients with NSCLC. Despite increasing evidence of the involvement of the TIGIT/PVR axis, few studies have addressed the prognostic role of PVR 22. Terefore, we aimed to assess PVR expression in surgically resected SQCC tissues to determine its relationship with PD-L1. We evaluated the correlations of PVR and PD-L1 expression with clin- icopathological factors and determined their prognostic implications for survival outcomes in SQCC (Fig. 1). Results PVR expression in SQCC (The Cancer Genome Atlas [TCGA] data). We analyzed TCGA RNA- sequencing data of gene expression profles in SQCC to identify the co-expression patterns of immune check- point ligands (ICLs)23. Figure 2a, b consists of two main clusters, including the PD-L1 (CD274) cluster of 10 other ICLs (CD48/80/86, BTN2A2/3A1, TIMD4, VSIG4, PDCD1LG2, TNFRSF14, and LGALS9) and the PVR cluster of PVR, NECTIN2, and CD276. Clustering analysis showed no correlation between the PD-L1 (CD274) cluster (red boxes) and the PVR cluster (blue boxes). Scatter plot also showed no correlation between the PD-L1 (CD274) cluster and the PVR cluster (Fig. 2c). According to the TCGA data, PVR was expressed independently of PD-L1 expression in SQCC. We also analyzed the survival diference for both high expression and low expression of PD-L1 and PVR but found no sig- nifcant diferences in OS between the groups with high and low PD-L1 (P = 0.91) and PVR expression (P = 0.55; Fig. 2d). We further categorized patients into four groups depending on PD-L1 and PVR expression: PD-L1hi/ PVRlo, PD-L1lo/PVRlo, PD-L1hi/PVRhi, and PD-L1lo/PVRhi (Fig. 2e). Similarly, no signifcant diferences were found between the four subgroups using multivariate analyses. Tere was a trend for the PD-L1hi/PVRhi group to have the worst survival rate, indicating that high expression of PD-L1 and PVR may contribute to poor prognosis. Baseline clinicopathological characteristics. In total, 259 patients were included (see Supplementary Table S1 online). Patients were distributed evenly across the age groups (< 65 years, 47%; ≥ 65 years, 53%). Most patients were male (93%, n = 241). When classifed according to American Joint Committee on Cancer (seventh edition) stage, the majority of patients were initially diagnosed with stage 1 disease (47%, n = 122), followed by Scientifc Reports | (2021) 11:8551 | https://doi.org/10.1038/s41598-021-87624-x 2 Vol:.(1234567890) www.nature.com/scientificreports/ Figure 2. Co-expression patterns of immune checkpoint ligands (ICLs) and immune checkpoint receptors in Te Cancer Genome Atlas data of lung squamous cell carcinoma. (a) Heatmap showing Spearman’s correlation coefcients for all pairs of ICLs. Te dendrogram shows the results of hierarchical clustering of ICLs based on the correlation coefcients using Euclidean distance as a dissimilarity measure and the single linkage method. Red boxes represent programmed death ligand-1 (PD-L1/CD274); blue boxes represent poliovirus receptor (PVR). (b) Gene expression patterns of ICLs among four patient groups: (1) PD-L1hi/PVRlo, (2) PD-L1lo/PVRlo, (3) PD-L1hi/PVRhi, and (4) PD-L1lo/PVRhi. Red represents increased median expression and blue represents decreased median expression levels of each ICL. Te colored bar shows the gradient of log 2 fold-changes in expression levels. (c) Scatter plot showing patterns of PD-L1 and PVR. (d) Kaplan–Meier analysis of overall survival according to high and low PD-L1 (CD274) and PVR expression, respectively. (e) Multivariate analyses of four patient groups show no statistically signifcant diferences between the groups. stage 2 (29%, n = 76) and stage 3 disease (21%, n = 54). Pleural involvement and lymphovascular invasion were observed in 26% (n = 68) and 11% (n = 28) of patients, respectively. Most patients had a history of smoking, with 41% (n = 106) being smokers at the time of hospitalization; 44% (n = 114) were former smokers. Adjuvant chemotherapy and radiotherapy were administered to 85% (n = 221) and 17% (n = 44) of patients, respectively.